CHAPTER 3
Quine on
Analyticity
Chapter
three contains five main topics, based on Quine’s
essay “Two Dogmas of Empiricism”. First
we are reminded that meaning is not naming, followed by a section which
examines definition and meaning. The
third section examines how analyticity is grounded in semantic rules, and why Quine abandons analyticity.
Section four considers the radical reductionism in empiricism, and the
problems with the theory of verification.
Finally the fifth section examines the conclusion to Quine’s
important essay “Two Dogmas of Empiricism”.
In order to
demonstrate how ontological relativity fails to explain what variables are
variables of, a discussion of Quine’s view of
epistemology is necessary. The rejection
of the analytic-synthetic distinction is discussed in Quine’s
essay “Two Dogmas of Empiricism”. The
two dogmas of empiricism are: (1) the
belief in an analytic-synthetic distinction, and (2) the dogma of reductionism, or the
belief that all statements can, in complete isolation from all the others,
admit the confirmation or infirmation of all the
others. [126] Quine defines analytic as “grounded in meanings independent
of fact”.[127]
Synthetic is the opposite or “grounded in fact”.[128] Quine defines the dogma of reductionism as: “the belief
that each meaningful statement is equivalent to some logical construct upon
terms which refer to immediate experience.”[129] The outcome of abandoning these dogmas of
empiricism, according to Quine is two-fold: (1)there will be a
blurring between metaphysics and the natural sciences, and (2) a shift toward
pragmatism. [130] Kant
was not the first to make the analytic-synthetic distinction. Leibniz had his distinction between truths of
reason and truths of fact.[131] Truths of reason were true in every possible
world, and not even God could change these truths. Quine defines
truths of reason as “those which could not be possibly false.”[132]
What verifies such truths is that they cannot violate the law of
contradiction. Nothing can both be and
not be at the same time and in the same respect. This is what an analytic statement is: a
statement that cannot be denied without contradiction. Quine
thinks this is a definition of very little value.[133] Self-contradictoryness
is a board notion that too needs clarification just as much as the notion of
analyticity does, according to Quine.[134] Truths of fact are verified by sense
experience, and are only true if observed or true in this actual world (the
best of all possible worlds). Synthetic judgments are judgments which cannot not be verified without sense experience to legitimate
them. Truth is a synthesis of the understanding of the meaning of the
terms used and verification of the states of affairs they describe as being the
case in reality. The truth of synthetic
statements is contingent upon reality, and analytic statements are necessarily
independent of reference to sense data, and are sometimes called tautological
since they are always true.
Hume called this distinction relations between ideas and matters of
fact.[135] This is sometimes called Hume’s fork. Relations between ideas are also verified by
the law of contradiction, and are necessary.
Only the relation between the ideas in the judgment
are examined. Matters of fact are
verified by sense experience, and hence are contingent upon being true in the
world or fact. Kant made the distinction
between the analytic (necessary, independent of sense experience) and the
synthetic (contingent, dependent upon sense experience). Descartes had too made such a distinction
before him. An analytic statement is a
statement in which the predicate term is contained in the subject term, or as Quine defines it: “one that attributes to its subject no
more than is already conceptually contained in the subject.” [136]
MEANING IS NOT NAMING
Quine reminds us that meaning is not naming.[137] There are several illustrations of the fact
that “terms can name the same thing but differ in meaning.” [138] Consider the following example. While driving down an unfamiliar road while a
passenger is giving directions, the driver may ask for clarification: “Turn left?”
The passenger responds by exclaiming “Right!” So what is the driver to
do; should the
driver turn right, or is the left turn in fact correct? Such semantic ambiguity is common among words
that are the same, but have different meanings. Quine is refering to another kind of ambiguity. An example is the age old confusion of the
morning star and the evening star as two separate objects, when in fact the
different terms refer to the same object, the planet Venus. Sure enough, the morning star is there in the
early morning, and the evening star in the evening, but to think that they are
two separate objects is the true confusion.
Being both the morning star and the evening star is true of the planet
Venus. General terms, or predicates,
whether they be concrete or abstract, such terms do
not name entities, but are “true of” an entity according to Quine.
[139] The example that “9” and
“the number of planets” name the same abstract thing, but are not at all
similar in meaning. [140] This can not be
solved using an analytic process, according to Quine, since empirical
evidence is needed to show that there are in fact 9 planets in our solar
system. “The class of all entities of
which a general term is true is called the extension.” [141]
Extension is the set of all things to which the general term applies. There is a parallel, Quine
continues, between the meaning of a singular term, and the entity it is to
name, and the meaning of a general term, and the class of all things to which
the general term applies (extension). [142] The confusion of meaning with general terms
is not as typical as confusion of meaning with singular terms.[143] It is, Quine
contends, typical to reject intension (connotation) for extension (denotation),
like Ockham had.[144] Quine notes how the
essentialism of Aristotle led to the modern notion of intension.[145] Aristotle thought that there were essential
qualities that made a thing the kind of thing that it is.[146] Accidental qualities in contrast are
qualities which are not necessary to make an thing a
thing that it is. Aristotle thought that
man was essentially a rational animal.
This essentialism of Aristotle is different from meaning Quine points out.[147] “Things have essences, for Aristotle, but
only linguistic forms have meanings.
Meaning is what essence becomes when it is divorced from the object of
reference and wedded to the word.”[148] This is no longer a question about things,
but of meanings. “[W]hat sort of things
are meanings?”[149] First Quine points
out that meaning and reference are distinct from one another.[150] The theory of meaning depends on the synonymy
of linguistic forms and the analyticity of statements. [151]
But, Quine sees no validity in synonymy nor analyticity, and hence he says that meanings ought to
all together be abandoned. [152] He examines the problem of analyticity.
Analytical
statements fall into two classes according to Quine; the first class is called “logically true”
and the second class of analytical statements depend on the notion of
“synonymy,” which Quine thinks also needs
clarification. [153] An example of a first class analytical
statement is, according to Quine, “No unmarried man is
married.” [154] This statement, Quine
contends, is logically true, or “remains true under all reinterpretations of
its components other than the logical particles.”[155]
The subject term contains a double negation and the predicate an affirmation of
the same term. He describes unmarried man in this way since he doubts the
synonymy of unmarried man with bachelor.
Hence, there is this other class, the second class,
and Quine uses the example following for an
analytical second class statement: “No bachelor is married.” [156] The second class uses synonyms, which are
logically containing the same information, but Quine
seems skeptical.[157] Some languages may lack synonym pairs, and
for such languages, Carnap appealed to what he called
“state-descriptions”.[158] In such languages “No bachelor is married”
is a synthetic statement.[159] “Analyticity serves its purpose only if the
atomic statements of the language are [...] mutually independent.”[160] If truth were assigned to both “John is
married” and “John is a bachelor” then “No bachelor is married” becomes a
synthetic statement. This example of the
second class depends heavily upon the notion of synonymy.[161] Such an example would violate the law of
non-contradiction if such qualities were essential to John’s being John. John cannot be both a bachelor and married at
the same time. Quine
is certain that “not not married” is synonymous with
“married”, but does not think that “not bachelor” is synonymous with “married”.
DEFINITION AND MEANING
Since
definitions rely on synonymy, Quine brings his
attention to them. The first class of
analytical statements relied upon definitions. Quine’s
first example is:
(1) “A bachelor is an
unmarried man.”
“A bachelor is an
unmarried man” is a definition of bachelor.
Hence an unmarried man is both a necessary condition (this is a
necessary condition for that if and only if that cannot be without this) and
sufficient condition (this is a sufficient condition for that if and only if
this alone can guarantee that) for being a bachelor. That is how definitions are supposed to work,
if the synonymy is real. “[R]eport of an observed synonymy cannot be taken as the ground
of the synonymy.”[162] Quine thinks that the criteria for
determining synonymy is not at all clear, but synonymy is nonetheless
granted in the conventional use of language.[163] These conventions are established by
usage. But such usage has already been
rejected for grounds for accepting the synonymy.
Explication
is not synonymy, since it aims to aid in the understanding of the definiendum, and is not just merely an outright
repetition. But no matter, most all
definitions, with rare exception, rely on prior
notions of synonymy.[164] Definitions did not help with the
understanding of the grounds for synonymy.
Hence, Quine leaves that,
and moves on to discuss
interchangeability.[165]
Interchangeability is the ability to exchange two linguistic forms for one
another without any change of truth
value. [166] Quine explains that Leibniz called this salva
veritate.[167] This in no way eliminates vagueness, since
the linguistic forms may be synonymous in their
vagueness.[168] Quine contends to
demonstrate that all synonyms are not interchangeable salva
veritate. He easily has many options to replace
with bachelor, but his example is
“‘Bachelor’ has less than ten
letters.”[169] It is
quite apparent that ‘bachelor’ and ‘less than ten letters’ is not going to be salva veritate. However Quine is
not playing fair. The lack
of interchangability
is due to his using denoting (is true of) instead of connotation.
Assuming
once again that analyticity is possible Quine offers
another example:
(2)
“All and only bachelors are unmarried men.”[170]
This is to be a criterion for a
cognitively synonymous statement, and it is salva veritate, in which the truth is saved. Quine is looking
for an account of analyticity which does not
depend on prior synonymy. Is this a
sufficient condition for cognitive synonymy, the interchangeability of
bachelors and unmarried men, as in the statement: “All and only
bachelors are unmarried men.”[171] If that will not convince, perhaps the next
example will finally shed light on the grounds of analyticity:
(3)
“Necessarily all and only bachelors are bachelors.”
“[S]upposing necessarily [is] so narrowly
construed as to be truly applicable only to analytic statements.”[172] The resulting sentence, assuming salva veritate
would be:
(4)
“Necessarily all and only bachelors are unmarried men.”[173]
To claim that “Necessarily all and only
bachelors are unmarried men” is true is to say
that:
(5)
“Necessarily all and only unmarried men are bachelors”
is true and this
would make “All and only bachelors are unmarried men” an analytic statement,
and that bachelors and unmarried men are synonymous.[174] But Quine
contends that a deception has gone on here.[175] I had a student in one of my logic classes claim that her boyfriend was not a bachelor
and he was an unmarried man. This seems
to
pose a problem for the analyticity of
bachelor with unmarried man. This
example is not really a fair use of the term bachelor, since bachelors are
indeed necessarily unmarried
men.
But what of the married man that acts as if he is a bachelor or lives
the life of the playboy. Does such a person make it impossible of there to be
certainty in the meanings
of our terms? Surely he is married; ask his wife if she thinks he is both
married and at the same time a bachelor.
He is just playing the part, and is not in fact truly a bachelor as
bachelor are.
The
above statement supposes we are working with a language rich enough to contain
the adverb ‘necessarily’, this adverb being so construed as to yield truth when
and only when applied to an analytical statement. But can we condone a language
which contains such an adverb? Does the
adverb really make sense?[176]
The only way to make sense of the
adverb ‘necessarily’ is to already understand the grounds for the notion of
analyticity, which have not as yet been determined to Quine’s
satisfaction. Quine
thinks that the above argument is circular.[177] To admit some prior grounds for synonymy is
to admit to a priori knowledge, something Quine
is not willing
here it appeal to. Quine gives an example of
an extensional language to make his point. He makes a language using one place
predicates (F, where Fx
means x is a man); many
place predicates (Gxy,
where Gxy means x loves y);
atomic sentences which can be composed of one or more variables (‘x’, ‘y’,
etc.), and atomic functions (‘not’, ‘and’,
‘or’, etc.).[178] “Now a language of this type is extensional,
in the sense: any two predicates which
agree extensionally (that is, are true of the same objects) are
interchangeable salva veritate.”[179]
In
such languages, there is no certainty that synonymy will hold for the type as desired,
and there is no reliance upon meaning when it comes to random matters of
fact.[180] Quine thinks that
appeal to cognitive synonymy was wrong, though it did explain the analyticity
of “All and only bachelors are unmarried men.”[181] Cognitive synonymy
also works for many-place and one-place
predicates.[182] Quine concludes
that singular terms may be said to be cognitively synonymous when it is a
statement of identity using
an equivalence (p = q).[183]
And finally, Quine abandons salva
veritate (rooted in de re necessity) for
an interchangeability of his own salva analyticitate (rooted in descriptive
phrases). [184]
THE GROUNDING OF ANALYTICITY IN
SEMANTIC RULES
Quine abandons the discussion of definition and meaning and
focuses on analyticity, since its grounding is at the source of the issue.
Analyticity
at first seemed most naturally definable by appear to a realm of meanings. On
refinement, the appeal to meanings gave way to an appeal to synonymy or
definition. But definition turned out to be a will-o-the-wisp, and synonymy
turned out to be best understood by dint of prior appeal to analyticity itself.
[185]
Quine questions if the statement
‘Everything green is extended’ is actually analytic. [186] This
is an example of a de re necessity, that to have color, an object must
be
extended. Quine
explains this confusion is not because he fails to understand the meaning of
green or extended, but he is not sure what analytic means. [187] Quine claims that
in
ordinary language it is difficult to always
differentiate analytic statements from synthetic statements because of the
vagueness of ordinary language. But artificial languages can be
constructed, which follow “semantical
rules” which are supposed to eliminate any confusion and vagueness. [188]
An example of what such a rule does is place conditions
upon statements S, in language L,
that require they be analytic.[189]
“[T]he problem is to make sense of this relation generally, that is, for
variable ‘S’ and ‘L’.” [190] What does it
mean that S is analytic for L,
and limiting the bounds of L does not eliminate the vagueness. Any such “semantic rule” which contains the
term ‘analytic’ is not
meaningful because it is not yet clear what
being analytic for any language means. “[T]he rules contain the word ‘analytic’
which we do not understand!” S is only analytic to
language L, but what is this ‘analytic
for’ relation? How are we to determine
the relation if analyticity is not understood?
Quine attempts to dump the word ‘analytic’ for
a term,
K. [191] Hence
this would be “S is only K to L”. But Quine even
acknowledges that this only will lead to an understanding of what the relation
of S is to K, but not what K-ness
is
in and of itself. Once again,
analyticity, according to Quine, had eluded
clarification.
Quine examines another class of “semantic rules” which
simply include what have been called analytic statements among the class of
truths. It does not imply
anything about interchangeability. Such a rule would specify what
characteristics statements must have in order to be classified as truths. Hence the statement is not
analytic because it is necessarily true, but
is a truth because it adheres to the “semantic rules of truths” and this makes
it analytic. [192]
Not
every true statement which say that the statements of some class are true can
count as a semantical rule-- otherwise all truths would be “analytic” in
the sense of being true according to semantic rules. Semantical rules
are distinguishable, apparently, only by the fact of appearing on a page under
the heading ‘Semantical Rules’; and this heading is itself then
meaningless.[193]
Although it appeared that the second
class of analytical statement which adhere to “semantical
rules of truth” did not at all shed light upon what analyticity means, because
the rules themselves are
meaningless. Quine
then compares semantic rules to postulates.[194] Semantical rules
are meaningful, if they are seen as postulates Quine
contends.
“[I]f conceived in a similarly relative spirit-- relative, this time, to
one or another particular enterprise of schooling unconversant
persons in sufficient conditions
for truth of statements of some natural or artificial language
L.”[195] If this was the case, Quine
informs, then no semantical
truth would exclude any other. ‘Brutus killed Caesar’
would have a different meaning entirely if
killed has another meaning, such as begat. [196] “Thus
one is tempted to suppose in general that the truth of a statement is somehow
analyzable into a linguistic component and a
factual component.”[197] This is from where the false acceptance of
the analytic-synthetic distinction arises, and then it is acceptable
to find that analytical statements are
factually null.[198] But Quine does not
think a clear boundary between analytical statements and synthetical
statements has been
demonstrated.
“That there is such a distinction to be drawn at all is an unempirical dogma
of the empiricists, a metaphysical article of faith.”[199] Why empiricists have
accepted this a priori notion, which Quine calls mere faith, is quite beyond him, an Quine is not willing to accept the distinction.
Since
Quine had set aside the question of meaning earlier,
he returns to it now. “[W]hat
[...] of the verification theory of meaning?” Quine
asks. [200] The verification
theory of meaning is “the meaning of a
statement is the method of empirically confirming or infirming it. An analytic statement is that limiting case
which is confirmed
no matter what.” [201] Hence, usually empirical evidence is used to
support or verify meaning, but with analytical statements, since they are true
regardless of empirical
evidence, are not verified in this way. In reference to synonomy,
it can now be said that two statements are synonymous is and only if they are
verified by empirical confirmation
or infirmation.[202] Quine notes that this is not a comparison of linguistic
forms but of statements, but from the concept of synonymy of statements, the
synonymy of other
linguistic forms could be similarly defined. [203] “So if the verification theory can be accepted
as an adequate account of statement synonymy, the notion of analyticity is
saved after all.” [204] But Quine does not
accept the theory of verification. What
set of empirical criteria demonstrates that two statements are in fact
synonymous? “What, in
other words, is the nature of the relation
between a statement and the experiences
which contribute to or detract from it confirmation?” [205] This leads question leads Quine
to
examine the nature of radical reductionism
in empiricism.
RADICAL REDUCTIONISM IN EMPIRICISM
AND THE PROBLEMS WITH THE
THEORY OF VERIFICATION
Radical
reductionism is the view that “[e]very meaningful statement is held to be translatable
into a statement (true or false) about immediate experience.” [206] This notion
of radical reductionism is something
that empiricists Locke and Hume held.
Locke and Hume thought that knowledge was acquired after sense
experience, and hence all simple
ideas could be traced back to simple
impressions as their source. Simple
impressions were fleeting immediate sense experiences, and simple ideas are
fainter copies of
impressions in imagination. Complex ideas are compounds of simple
ideas. Hume found that he could
legitimate no ideas in this manner, since the bundles of perception
prevented him from tracing simple ideas back
to their simple impressions as their source. Hume realized the difficulty with
such a reductionism. Quine
acknowledges this by
saying:
“the doctrine remains ambiguous as between sense data and sensory events
and sense data as sensory qualities; it remains vague as to the admissible ways
of
compounding.” [207]
Compounds are made out of simples, but it quite possible to create compounds of
simples that have nothing to do with reality.
Since ideas are fainter copies
of impressions in imagination, it
cannot be determined if any of our ideas are really correct in describing the
world. The true problem of the theory of
verification is in its
blind acceptance of such an reductionism,
since such empiricists only try to match words to sense-datum, and not the
entire sentence in question. This is a
type of reductionism
that Quine does
not find tolerable. Quine
wants the entire sentence to be translated into sense-datum language. [208] Quine fails to recognize that descriptions do not solve the
problem.
Quine thinks that Locke and Hume would welcome
this acceptance of the sentences as the significant units rather than the
words. [209] “This reorientation whereby
the primary vehicle of meaning came to
be seen no longer in the term but in the statement.” [210]
This is an important shift in the verification of meaning, and now it is the
sentences which must be verified by sense
experience. Carnap
tried to translate significant discourse into a sense-datum verified language. [211] Carnap did not start with a
sense datum language, because his language
included notions of logic and set theory. [212] Carnap was able to
define some sensory concepts with his system, and: “[h]e was the
first empiricist who, not content with
asserting the reducibility of science to terms of immediate experience, took
serious steps toward carrying out the reduction.”[213]
Carnap had not seen
that such system, even of the most basic statements, could not clearly show
that they were true about the world.[214]
Because of this vagueness, Carnap’s
theory cannot be used to show that
our ideas do match the world, but Carnap did show that
it is statements, and not words that have to face up to the scrutiny of
empirical
verification.
THE TWO DOGMAS OF EMPIRICISM
The
two dogmas of empiricism: (1) the
analytic synthetical distinction, and (2) the dogma
of reductionism, are in fact two aspects of the same dogma, since at root the
dogmas are identical. [215]
We
lately reflected that in general the truth of statements does obviously depend
both upon language and upon extralinguistic fact; and
we noted that this obvious circumstance carries in its train, not logically but
all too naturally, a feeling that the truth of a statement is somehow analyzable
into a linguistic component and a factual component. The factual component must, if we are
empiricists, boil down to a range of confirmatory experiences. In the extreme
case where the linguistic component is all that matters, a true statement is
analytic.[216]
Quine thinks that true empiricists would
require the verification of statements using sense-data. But there is still this so-called false
notion of analyticity and syntheticity of
statements according
to empiricists, but even so, there still seems to be the existence of
statements that do not require sense-data for verification, but can be verified
using the linguistic component
alone.
So it seems that Quine wants to eliminate the analytic-synthetic
distinction, but still insists on making the verified by sense-data or verified
by linguistic
component distinction that is at the root of
the analytic-synthetic distinction. I have long thought that there was a false dualism regarding the
analytic-synthetic distinction.
But empiricists tend to favor the
sense-datum verification, while rationalists favor the linguistical
verification, neither is willing to admit that the other form of verification
does not exist or is
not meaningful. Hence there are always bound to be linguistically verified
statements in any empirical theory, as well as empirically verified statements
in any
rationalistic
theory. Quine
hopes that his discussion has shown that the distinction between the analytic
and the synthetic has been shown to be a distinction which is hard to make in a
clear, distinct way.[217] Quine thinks the
real problem is in thinking that there is this division between the linguistical component and the factual component of
individual statements. [218]
Science is dependent on both language
and experience and it is this belief of some dualistic nature of statements has
made some favor the empirical rather than the linguistic component,
since science has statements that seem to
not reflect this duality, science may think that it is free of the dilemma. [219] Quine notes that words were too narrow for verification,
and so are mere
statements.
“The unit of empirical significance is the whole of science.” But we
must bracket in some way to make verifications.
Postulates are a form of convention.
Logic is
must rest on a convention. But this convention cannot be merely
definitional, since definitions do not found truths, only transform them. Quine would not
accept a type of
conventionalization of analytical definitions, that make
them merely conventionally necessarily true.[220]
Quine defines science as being like “a field of force whose
boundary conditions are experience.” [221] Science is bound with sense-datum
verification. There has to be a
method of determining truth value, and such
values range over statements. Quine takes a pragmatic approach to truth in “Two Dogmas of
Empiricism”. “Truth values have to
redistributed over some of our statements. Reevaluation of some statements entails reevaluation
of others, because their logical interconnections-- the logical laws being in
turn simply certain further statements of
the system, certain further elements of the field.”[222]
But experience is not bound by determined boundaries, and it is difficult to
know what to reevaluate and in light of
what evidence. It is clear that
particular statements cannot be traced back to particular experiences, in the
same way that simple ideas could
not be traced back to simple
impressions. All we can do is to take a
pragmatic approach,
and “accept indirectly through considerations of equilibrium
affecting the field as a whole.”[223]
Hence all the statements make a
field, and each statement is evaluated, and
reevaluated again, and is judged according as it best fits into our system of
beliefs. No statement is immune
to revision. [224] And what of the abstract
entities of mathematics?
“Epistemologically these are myths on
the same footing with physical objects and gods, neither better nor worse
except for differences in degree to which they expedite our
dealings with sense experiences.” [225]
Hence ontological concerns are on the same level as questions of science and
mathematics. It is important to the
natural sciences to find a
conceptual framework. The only real way to make a distinction
between ontological concerns and scientific concerns is by maintaining the
distinction between analytic and
synthetic.
Quine wants to eliminate such a distinction it
seems without eliminating the factual component of
verification nor the linguistic component of verification. What he
is not in favor of is a reductionism of
one kind of statements to another. So
how can this factual-linguistic distinction be eliminated, or at least, how can
there be two ways of
verification if there is not a analytic-synthetic
distinction? Linguistically verified statements
are not required to be verified by sense experience, and hence the dichotomy
cannot be neatly reduced to a singular type
of verification. Perhaps understanding the us of
predication, the possibility of reification may be made more
clear.
[131] Monadology #33, 258.
[137] Also see sections 37, 41, and 42 in Word and Object for more about this point. Although Quine abandons singular terms and names here, he reparsed names later in his career.
[147] “Two Dogmas Of Empiricism,” 22. Quine is mistaken about Aristotle’s view. According to Aristotle, the subject was part of what ousia (what it was to be or quid erat esse) means, and this entails using linguistic forms that have meanings. The formal cause includes the definition (as well as the shape and form). Metaphysics 1010 A - 1013 A, and Categories 5.
[149] Ibid. Meanings are things. Linguistic forms have essences. Quine is in fact setting up a kind of essentialism much like Aristotle’s (see Posterior Analytics).
[158] Ibid.
[160] Ibid. Or, they are not defined in terms of each other. This is close to Descartes’ definition of substance, as that which can be conceived alone.
[161] “Two Dogmas Of Empiricism,” 24.
[163] “Two Dogmas Of Empiricism,” 25.
[164] “Two Dogmas Of Empiricism,” 27.
[166] “Two Dogmas Of Empiricism,” 27.
[167] Ibid.
[169] “Two Dogmas Of Empiricism,” 28.
[170] “Two Dogmas Of Empiricism,” 29.
[177] The argument is circular. “Two Dogmas Of Empiricism,” 30.
[189] “Two Dogmas Of Empiricism,” 33.
[192] “Two Dogmas Of Empiricism,” 34. Quine has totally missed the point here. He acknowledges that semantic rules are merely arbitrary, but thinks that necessity, if there is any, is based on adherence to such rules, and this is where the notion of analyticity comes from. The semanitic rules require necessity, hence he has put the cart before the horse. An example of such a rule is that analyticity requires that the predicate term of a proposition be contained with its subject. Quine needs to attack de dicto necessity, and not merely just de re necessitiy.
[202] Ibid.
[210] Ibid.
[223] Ibid. Quine abandons the pragmatic approach for the coherence approach in Word and Object and Web of Belief. For more about this see Word and Object, page 23, and Pursuit of Truth pages 33, 79, 134-135.
[224] Ibid.